Prevalence and impact of hepatitis B and C virus co-infections in antiretroviral treatment naïve patients with HIV infection at a major treatment center in Ghana
Abstract
Data on the effects of the presence of hepatitis B virus (HBV) and hepatitis C virus (HCV) in patients co-infected with these viruses and HIV in West Africa are conflicting and little information is available in Ghana. A cohort of 138 treatment naïve individuals infected with HIV was screened for HBV and HCV serologic markers; HBsAg positive patients were tested for HBeAg, anti-HBe, and anti-HBc IgM. The viral load of HIV-1 in the plasma was determined in 81 patients. Eighteen of the 138 patients (13%) and 5 (3.6%) had HBsAg and anti-HCV, respectively. None of the patients had anti-HBc IgM, but 10 (55.6%) and 8 (44.4%) of the 18 patients who were HBsAg positive had HBeAg and anti-HBe, respectively. In patients with measurement of CD4+ undertaken within 1 month (n = 83), CD4+ count was significantly lower in patients with HBeAg (median [IQR], 81 [22–144]) as compared to those with anti-HBe (median [IQR], 210 [197–222]) (P = 0.002, CI: −96.46 to 51.21). However, those with HIV mono-infection had similar CD4+ counts (median [IQR], 57 [14–159]) compared to those with HBeAg (P = 1.0, CI: −71.75 to 73.66). Similar results were obtained if CD4+ count was measured within 2 months prior to initiation of HAART (n = 119). Generally, HBV and anti-HCV did not affect CD4+ and viral loads of HIV-1 in plasma but patients with HIV and HBV co-infection who had HBeAg had more severe immune suppression as compared to those with anti-HBe. This may have implication for initiating HAART in HBV endemic areas. J. Med. Virol. 84:6–10, 2011. © 2011 Wiley Periodicals, Inc.
INTRODUCTION
Data on individuals co-infected with HIV and hepatitis B virus (HBV) in West Africa is still emerging [Diop-Ndiaye et al., 2008; Adewole et al., 2009; Idoko et al., 2009; Laurent et al., 2009], but this is not limited to this region alone [Harania et al., 2008; Omland et al., 2008].
While some studies have suggested that the presence of HBV co-infection is associated with a more severe depression of pre-HAART CD4+ counts as compared to those with HIV infection alone [Diop-Ndiaye et al., 2008; Omland et al., 2008; Otegbayo et al., 2008; Adewole et al., 2009; Idoko et al., 2009], other investigations found no differences [Firnhaber et al., 2008; Harania et al., 2008; Laurent et al., 2009]. Discrepant findings have also been reported on the effects of HBV co-infection on measurements of pre-HAART viral load of HIV-1 in plasma (pre-HAART HIV-1 VL). In a European cohort, the pre-HAART HIV-1 VL was slightly higher in individuals co-infected with HIV and HBV as compared to those with HIV mono-infection [Konopnicki et al., 2005]. This observation was more pronounced in a Nigerian cohort with HIV and HBV co-infection [Idoko et al., 2009]. Several other studies with patients co-infected with HIV and HBV or with HIV and the presence of anti-hepatitis C virus (HCV), did not show any differences in pre-HAART HIV-1 VL as compared to those with HIV infection alone [Diop-Ndiaye et al., 2008; Harania et al., 2008; Omland et al., 2008; Otegbayo et al., 2008; Laurent et al., 2009].
Generally, the presence of anti-HCV in patients infected with HIV is low in West Africa [Diop-Ndiaye et al., 2008; Harania et al., 2008; Otegbayo et al., 2008; Adewole et al., 2009]. The challenge has therefore been on improving treatment regimens for individuals with HIV and HBV co-infection when more patients have access to antiretroviral drugs [Firnhaber et al., 2008; Harania et al., 2008; Laurent et al., 2009].
Few studies have examined the prevalence and effects of HBV and HCV infections in individuals infected with HIV in Ghana [Brandful et al., 1999]. Currently, no reliable cohort data are available on the effects of HBV or anti-HCV on pre-HAART CD4+ and pre-HAART HIV-1 VL. With the increasing use of antiretroviral therapy (ART), there is the need to determine the effects of HBV and HCV infections on baseline parameters in individuals infected with HIV initiating HAART in order to interpret short- and long-term outcomes. This study therefore sought to determine the prevalence of hepatitis B surface antigen (HBsAg) and anti-HCV in patients naïve to ART and infected with HIV. The demographic characteristics, pre-HAART CD4+ and pre-HAART HIV-1 VL were compared for the co-infected and HIV mono-infected groups.
PATIENTS AND METHODS
Patient Population
The study was performed at the Fever's Unit of the Korle Bu Teaching Hospital, Accra, Ghana, from June to November 2007. HIV infected individuals with CD4+ counts ≤250 cells/µl who were at least 18 years old, antiretroviral drug naïve, and about to initiate HAART were eligible for the study. Using the formula for proportions n = z2 (p) (1 − p)/e2, where z is the standard score for the confidence for the confidence level (1.96), p is the sample proportion; e is the allowable error (5%) and prevalence of 8% for HBsAg (Lartey, 2005; personal communication), a minimum of 113 patients were necessary for the study. Clinical data were obtained from patient folders, and CD4+ count determination done at study entry or during previous patient visits with the FAScount (Becton-Dickinson, Franklin Lakes, NJ).
The study was approved by the University of Ghana Medical School Ethics and Protocol Review Committee and informed consent was obtained from patients before being enrolled in the study.
Blood Processing and Virological Tests
Whole blood from patients was obtained using EDTA tubes and samples transported to the laboratory on ice packs or kept at +2 to +8°C. Blood was centrifuged within 6 hr after collection using the EBA 20 table-top centrifuge (HettichZentrifugen, Tuttlingen, Germany) and plasma kept in aliquots of approximately 600 µl at −85°C during long-term storage.
Plasma from samples obtained at enrollment were screened for HBsAg using a third generation ELISA, Surase B 96 (TMB)(General Biologicals Corp., Hsin Chu, Taiwan) with the Core HBsAg rapid test (Core Diagnostics, Birmingham, UK) used as a supplemental test. All plasma samples positive for HBsAg were tested for hepatitis B e antigen/antibody (HBeAg/anti-HBe) using the EASE BN-96 (TMB), and anti-HBc IgM (IgM antibodies for hepatitis B core antigen) using the ANTICORASE MB-96 (TMB) kits, respectively (General Biologicals Corp.). Anti-HCV was detected with SP-Nanbase C-96, also from General Biologicals Corp., and the Foresight HCV Antibody EIA (ACON Laboratories Inc., San Diego, CA) was used for supplemental testing. Pre-HAART HIV-1 VL determinations were performed on plasma samples from 81 patients with the COBAS Amplicor Monitor v1.5 and the COBAS Amplicor analyzer (Roche Diagnostics GmbH, Mannheim, Germany). These plasma samples were from all patients co-infected with HBV or having anti-HCV and a cross-section of patients with mono-infection conveniently selected.
Analysis of Data
The variables considered in the study include age, gender, pre-HAART CD4+ counts, pre-HAART HIV-1 VL, duration between the last CD4+ measurement and the initiation of HAART, World Health Organization (WHO) clinical disease staging, HBV infection, anti-HCV status and the “e” antigen status of patients with HBV infection. ANOVA was used in comparing age, pre-HAART CD4+ count, and pre-HAART HIV-1 VL of patients with HIV infection alone to those co-infected with HBV or exposed to HCV indicated by the presence of anti-HCV. The χ2-test was used to determine differences in gender and WHO clinical disease staging. The Tamhane's post hoc analysis was performed to identify groups that had significantly different pre-HAART CD4+ counts. The SPSS Version 14 software (SPSS, Inc., Chicago, IL) was used for statistical analysis.
RESULTS
Patients Characteristics and Hepatitis B and C Markers
A total of 154 patients were enrolled in the study, but 16 were disqualified due to incomplete clinical data. Of the remaining 138 patients, 40 (29%) were male, 98 (71%) were female, and 137 (99.3%) had their WHO clinical disease staging available. There were fewer men than women (1:2.45), and women (median age, IQR; 37.4 years) were more likely to be younger than men (median age, IQR; 37.4 years), P ≤ 0.001 [CI: 4.303–10.842]. Pre-HAART CD4+ count, pre-HAART HIV-1 VL, and the WHO clinical disease staging were similar for both women and men (P ≥ 0.05). For those with pre-HAART HIV-1 VL results, only one patient may have had an HIV-2 infection as indicated by the computerized report from the COBAS Amplicor and so excluded from the analysis.
Eighteen (13%) patients had HBsAg, and five had (3.6%) anti-HCV. No patient was found to be both HBsAg and anti-HCV positive. Of the 18 HIV patients infected with HBV, 10 (55.6%) carried the HBeAg, while 8 (44.4%) had anti-HBe. None of the 18 patients co-infected with HIV and HBV had anti-HBc IgM.
Baseline Parameters and Co-Infection With Hepatitis Viruses
The presence of HBV infection or anti-HCV was not associated with gender (P = 0.615), and did not affect the WHO clinical disease staging of patients (Table I). Nevertheless, there was a possibility of patients co-infected with HBV or having anti-HCV having different WHO stages as compared to those with HIV mono-infection (P = 0.086) (Table I). Age was not a risk factor for co-infection with HBV or HCV (P = 0.396), when HBV infection and the presence of anti-HCV were considered (P = 0.490), or when the HBeAg or anti-HBe status in HBV infections were considered (P = 0.564). The presence of HBsAg or anti-HCV presence did not affect pre-HAART CD4+ count when all the 138 patients were included in the analysis (Table I). However, when the HBeAg/anti-HBe status of patients co-infected with HBV was taken into account, patients with HBeAg appeared to have lower CD4+ counts (P = 0.059). This trend was shown to be significant when CD4+ counts measured within specific time periods preceding initiation of HAART was analyzed. CD4+ cells counts measured within 1 month before initiation of HAART (n = 83), were significantly lower in patients in whom HBeAg was also present (median [IQR], 81 [22–144]), compared to patients who showed only anti-HBe (median [IQR], 210 [197–222]) (P = 0.002, CI: −96.46 to 51.21). Those with HIV mono-infection had similar pre-HAART CD4+ counts (median [IQR], 57 [14–159]) compared to those with carrying HBeAg (P = 1.0, CI: −71.75 to 73.66). Similar results were obtained when all patients whose CD4+ had been done within two months before initiating HAART was analyzed (n = 119; P = 0.03).
HIV alone (n = 115) | HIV with HBsAg or anti-HCV (n = 23) | P-value* | HIV with HBsAg (n = 18) | P-value* | |
---|---|---|---|---|---|
Women | 83 (72.2%) | 15 (65.2%) | 0.396 | 11 (61.1%) | 0.405 |
Age (years); median (IQR) | 32 (32–47) | 39 (34–43) | 0.182 | 37 (34–42) | 0.490 |
WHO disease stage | 0.086 | 0.169 | |||
1 | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | ||
2 | 3 (2.6%) | 2 (8.7%) | 2 (11.1%) | ||
3 | 88 (76.5%) | 20 (87.0%) | 15 (83.3%) | ||
4 | 23 (20.0%) | 1 (4.3%) | 1 (5.6%) | ||
n.a. | 1 (0.9%) | 0 (0.0%) | 0 (0%) | ||
Ranked CD4a (cells/µl); n/[median, IQR] | 90 (17–168) | 124 (55–195) | 0.182 | 137 (35–197) | 0.246 |
1 | 66/57 (14–159) | 17/136 (60–199) | 0.075 | 13/137 (32–199) | 0.120 |
2 | 49/(22–184) | 6/(42–193) | 0.924 | 5/124 (28–207) | 0.925 |
HIV-1 pVL+ (log10 copies/ml); n/[median, IQR] | 58/5.49 (5.23–5.97) | 22/5.34 (5.10–5.79) | 0.146 | 17/5.4 (5.15–5.74) | 0.275 |
- a Ranked CD4, analysis considered pre-HAART CD4 that had been done within 1 month or after 1 month.
- * P values were determined by ANOVA and Chi-squared test and a value of <0.05 was considered significant.
DISCUSSION
In the same geographical setting, different studies have shown different effects of HBV infection on outcomes in patients with HIV infections [Otegbayo et al., 2008; Adewole et al., 2009; Idoko et al., 2009]. There is therefore the need for every clinic providing ART to determine the pre-HAART effects of HBV infections on pre-HAART characteristics of patients infected with HIV before initiating ART. This will help to evaluate outcomes after short- and long-term therapy.
The prevalence of HIV and HBV or HCV co-infections in this study is similar to others previously reported in West Africa [Diop-Ndiaye et al., 2008; Otegbayo et al., 2008; Adewole et al., 2009; Idoko et al., 2009] and confirms that HIV/HBV/HCV triple infection is not common [Diop-Ndiaye et al., 2008; Otegbayo et al., 2008]. Furthermore, the absence of anti-HBc IgM in the study participants indicates that their HBV infections were longstanding infections. Since HCV viral load was not determined in this study its influence could not be inferred.
The results of this study are consistent with previous observations of the high prevalence of HBeAg in individuals with HIV and HBV co-infection [Lesi et al., 2007; Miailhes et al., 2007; Osborn et al., 2007; Omland et al., 2008; Rouet et al., 2008; Idoko et al., 2009]. This also supports the report that HIV infections increase HBV replication in vivo resulting in high HBeAg titers [Goldin et al., 1990]. Since the patients with HBeAg were more immune-suppressed than those with anti-HBe, it is possible that pre-HAART CD4+ counts may have to reach a critically low level for the re-emergence and persistence of HBeAg [Gilson et al., 1997].
In HBsAg negative individuals, CD4+ levels have been shown to be higher than in individuals with HBV infections [Ercilla et al., 1986; Novick et al., 1986; Mazzetti et al., 1987]. Furthermore, the presence of HBeAg has been shown to be an indicator for low pre-HAART CD4+ in a Nigerian cohort [Idoko et al., 2009]. Thus, the low pre-HAART CD4+ in HBeAg positive patients may have been facilitated by the replication of HBV. The results of this study are therefore consistent with the aforementioned study conducted in Nigeria which had a comparatively larger patient population and wider pre-HAART CD4+ range.
Similar pre-HAART CD4+ values were seen in HIV-infected patients with anti-HCV and those co-infected HBV in the presence of HBeAg. The similarity between these two groups and those with HIV alone may be accounted for by other factors such as a comparatively larger group sizes and opportunistic infections (OIs) [Gallant et al., 1994]. Since the presence of anti-HCV may result in more severe immune-suppression in individuals co-infected with HIV and HCV [Greub et al., 2000; Sulkowski et al., 2002], the pre-HAART CD4+ count in patients with anti-HCV may reflect a true observation.
As has been suggested for HIV and HCV co-infections [Nunez et al., 2006; Korner et al., 2009], the presence of HBeAg or anti-HCV in HIV-infected patients with HBV or possible HCV co-infection may serve as a catalyst for apoptosis, thereby accelerating the depletion of CD4+ cells before the onset of therapy. Indeed, a recent study has shown that the presence of occult HBV infection in persons infected with HIV is associated with lower CD4+ counts [Cohen Stuart et al., 2009]. Since individuals co-infected with HIV and HBV may increase substantially when considering occult infections [Firnhaber et al., 2009], this may partly account for the similar pre-HAART CD4+ counts of the patients with HBeAg as compared to those with HIV mono-infections.
The reports of similar pre-HAART HIV-1 VL in individuals co-infected with HIV and HBV and those with HIV mono-infections has been observed in some studies [Harania et al., 2008; Laurent et al., 2009], but not in others [Konopnicki et al., 2005; Idoko et al., 2009]. Also, gender was not significant in predicting individuals infected with HIV who were likely to have HBsAg or anti-HCV, an observation which in contrast to other studies where men were shown to be more likely to have HBsAg [Diop-Ndiaye et al., 2008; Harania et al., 2008]. These differences may be difficult to explain in an African setting where data on injection drug users is not readily available.
In conclusion, the high prevalence of HBV infections among persons infected with HIV individuals suggests that screening for HBsAg should be routinely done before initiation of HAART and warrants the review of the treatment guidelines. Further studies are needed to understand the evolution of HBeAg and anti-HBe in African populations with HIV infection and its relation to CD4+.
Acknowledgements
We are grateful to the nurses and doctors at the Fevers Unit, Korle-Bu Teaching Hospital for making the study possible. Messrs Kwabena Obeng Duedu, Ernest Kwarteng, and Ms. Cecila Turkson are acknowledged for technical support and data collection. David Nana Adjei and Rev. Tom Ndanu of the College of Health Sciences, University of Ghana Medical School are also acknowledged for statistical support. The critical grammatical review of the manuscript by Dr. Diana Da-Ann Huang of Rush-Presbyterian-St. Luke's Medical Center, Chicago, Illinois, is acknowledged. HIV-1 plasma viral load kits were provided by the Programme Manager, National AIDS/STI Control Programme, and the University of Ghana staff PhD research grant.